CN105945311A - Numerically-controlled machine tool feed system speed regulation method based on power prediction - Google Patents

Abstract

The invention relates to a numerically-controlled machine tool feed system speed regulation method based on power prediction. During a rough turning process of a numerically-controlled machine tool, the minimum feed speed is generally set according to the maximum cutting depth, and machining is performed by taking the minimum feed speed as a constant feed speed. However, the problem of low efficiency caused by constant-speed machining is especially acute in case of a large amount of rough turning operation and a large machining batch. Aiming at the problem, the invention provides the numerically-controlled machine tool feed system speed regulation method based on power prediction. The method comprises the following steps: Step 1, acquiring power signals and inputting the power signals to a neural network to predict a power value of next moment; Step 2, calculating a deviation ratio of preset power and predicted power to judge whether speed regulation is needed; and Step 3, predicting a back cutting depth, and finishing speed regulation control over a feed system through combination of the back cutting depth and a preset power value. The method effectively solves the problem of lagging of speed regulation under different cutting conditions, reduces the machining cost and greatly improves the utilization rate and cutting efficiency of the machine tool.

Description

A speed regulation method of CNC machine tool feed system based on power prediction

technical field

The invention is suitable for use in the field of numerical control machine tool processing, and specifically relates to a speed regulation method for a feed system of a numerical control machine tool based on power prediction.

Background technique

Rough turning is a process of simple processing or primary processing of blanks. Its main purpose is to cut off the blank allowance and make the blank into approximate parts. Rough turning does not have high requirements on the quality of the processed surface, and the processing efficiency is the focus of attention. Therefore, in rough machining, the feed rate and depth of cut are generally set as large as possible, so as to remove as many chips as possible in the shortest possible time. For ordinary machine tools, rough turning is mainly done by the operator based on the work process and experience. However, for CNC machine tools, the machining process and process parameters work according to the program setting values. Once the programming is completed, the CNC machine tool will cut at a constant feed rate according to the set process parameters during the cutting operation. In the process of programming, in order to avoid damage to the tool, workpiece and machine tool due to cutting overload, the processing parameters must be selected according to the working conditions of the upper limit range of the load. However, in the entire cutting operation, the cutting amount is at the upper limit of the cutting depth, which generally only accounts for about 5% of the entire process, while the feed rate is always working according to the program setting value, which greatly reduces the processing time. efficiency. In the case of many roughing operations and large processing batches, the problem of low roughing efficiency is particularly prominent.

The core problem of controlling the feed rate is to analyze and control the cutting feed force. There are mainly two methods to obtain the feed force: off-line simulation analysis and on-line detection. The main problem of offline simulation analysis is that it cannot dynamically respond to changes in actual machining conditions. The simulation analysis results need to be corrected in combination with the actual working conditions, and the correction process is relatively complicated. By optimizing and controlling the real-time feed force, the feed speed can be controlled so that it changes with the actual cutting conditions, which has great advantages in realizing efficient cutting and stable operation of CNC machine tools. The real-time feed force can be measured by special force measuring instruments, and resistance strain force measuring instruments are widely used at present. However, this measurement method is inconvenient to measure, and it will affect the process system, weaken the rigidity of the machine tool, and cause difficulties in the installation of workpieces and tools, and the price is relatively high. On the other hand, because the speed regulation process of the feed system of the machine tool is carried out by the mechanical transmission mechanism, the response time is relatively long. Therefore, using the conventional feedback control method is prone to overload impact due to speed regulation lag, and damage to tools, components and machine tools. Therefore, it is suitable to combine the predictive control method with the speed regulation of the feed system.

According to the energy balance equation and energy consumption characteristics of the machine tool, the energy consumption can quickly respond to load changes. For the feed system of CNC machine tools, although there are many energy-consuming links in the system and the energy flow is complex, it is found that there is a quadratic function relationship between the energy input of the feed system and the feed force. At the same time, power changes during the roughing process of the machine tool are easy to detect and identify. In view of this, this patent proposes a speed control method for the feeding system of CNC machine tools based on power prediction to realize the speed control of the feeding system.

Contents of the invention

The purpose of the present invention is to provide a speed regulation method for the feed system of a CNC machine tool based on power prediction, which is aimed at the low efficiency of the roughing process under the condition of constant low-speed feed speed processing, and the response time of the speed regulation process of the machine tool feed system is relatively short. Long and poor real-time performance. By predicting the power at the next moment under the actual working conditions, establishing the functional relationship between the power and the amount of back cutting, changing the machining method of cutting at a constant feed speed during the rough turning process of the CNC machine tool, and realizing the optimization of the feed system during the rough turning process of the machine tool The optimization of speed regulation not only realizes the control of feed speed, but also effectively solves the problem of speed regulation lag, and improves the utilization rate of machine tools and cutting efficiency.

To achieve the above object, the present invention provides the following technical solutions:

A speed regulation method for a feed system of a CNC machine tool based on power prediction, comprising the following steps:

Step 1: Predict the power value at the next moment:

A voltage sensor, a current sensor and a power sensor are respectively installed to collect the voltage, current and power values at the input end of the feed motor. Filter the collected data and process the singular values; then calculate the collected voltage and current values, and fuse the calculated power value with the detected power value to improve the power detection accuracy.

The power flow of the servo motor in the feed system includes stator copper loss, iron loss, mechanical loss, stray loss, and the output power of the driving load. Combined with the steady-state voltage equation of the AC servo motor, the power of the servo motor can be expressed as:

${\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 3</span>}\mathrm{<span\; class="notranslate">\; R</span>}{\mathrm{<span\; class="notranslate">\; I</span>}}_{\mathrm{<span\; class="notranslate">\; the\; s</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span>\frac{{\mathrm{<span\; class="notranslate">\; \&omega;</span>}}_{\mathrm{<span\; class="notranslate">\; e</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&psi;</span>}}_{\mathrm{<span\; class="notranslate">\; d</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; \&psi;</span>}}_{\mathrm{<span\; class="notranslate">\; q</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; R</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; \&omega;</span>}}_{\mathrm{<span\; class="notranslate">\; e</span>}}{\mathrm{<span\; class="notranslate">\; K</span>}}_{\mathrm{<span\; class="notranslate">\; e</span>}}{\mathrm{<span\; class="notranslate">\; i</span>}}_{\mathrm{<span\; class="notranslate">\; q</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>$

Among them, R is the stator winding resistance; I _s is the effective value of the stator phase current; _{ω e} is the angular velocity of the electromagnetic field of the motor; ψ _d is the direct axis component of the magnetic flux; ψ _q is the quadrature axis component of the magnetic flux; K _e is the electromagnetic torque coefficient; is the quadrature axis component of the stator current. Combined with engineering practice, the power equation of the feed system can be obtained from formula (1):

$\begin{array}{c}{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 3</span>}{{\mathrm{<span\; class="notranslate">\; RB</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; K</span>}}_{\mathrm{<span\; class="notranslate">\; T</span>}}{\mathrm{<span\; class="notranslate">\; B</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; \&omega;</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 6</span>}{\mathrm{<span\; class="notranslate">\; RB</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; K</span>}}_{\mathrm{<span\; class="notranslate">\; T</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&CenterDot;</span><span\; class="notranslate">\; \&lsqb;</span>{\mathrm{<span\; class="notranslate">\; K</span>}}_{\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; q</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; m</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; m</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}\mathrm{<span\; class="notranslate">\; o</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; d</span>}}<span\; class="notranslate">\; )</span>\\ <span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; T</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; T</span>}}_{\mathrm{<span\; class="notranslate">\; c</span>}}<span\; class="notranslate">\; \&rsqb;</span>{\mathrm{<span\; class="notranslate">\; \&omega;</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 3</span>}\mathrm{<span\; class="notranslate">\; R</span>}{<span\; class="notranslate">\; \&lsqb;</span>{\mathrm{<span\; class="notranslate">\; K</span>}}_{\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; q</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; m</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; m</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}\mathrm{<span\; class="notranslate">\; o</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; d</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; T</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; T</span>}}_{\mathrm{<span\; class="notranslate">\; c</span>}}<span\; class="notranslate">\; \&rsqb;</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}\end{array}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>$

in, K _T is the torque coefficient; B _m is the motor damping coefficient; P is the pitch of the ball screw; μ _v is the viscous friction coefficient; K _g is the transmission ratio of the coupling; ω _m is the shaft speed of the feed motor; μ _c is Coulomb friction coefficient; M _t is the mass of the workbench; M _load is the mass of the workpiece loaded on the workbench; T ₀ is the internal loss torque of the motor; F _ext is the force on the workbench.

Take the processed power value as input, construct the training sample (x _k , y), input it into the neural network, and obtain the output value y, which is the predicted value of the power at time k+1. The power prediction refers to estimating the power value P _m+h (h>0) at the time t _m+h based on the detected power data (P _m , P _m-1 , . . . ) before the time t _m . The realization of this process is to perform nonlinear function fitting on the historical data (P _m ,P _m-1 ,…,) to generate a time-varying function P _m (t) about the variable t, and P _m+1 is related to the function P There is a mapping relationship between _m (t), that is, P _m+1 =F[P _m (t)]. Therefore, by fitting the functional F[·], the value of P _m+h (h>0) can be estimated.

Step 2: Calculate the deviation rate between the predicted power and the preset power, and judge whether speed regulation is required:

By obtaining the predicted power value, the speed regulation equation of the feed system in the rough turning process of CNC machine tools based on power prediction can be obtained as:

$\frac{{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; exp</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; x</span>}}}{{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; exp</span>}}}<span\; class="notranslate">\; \&le;</span>\mathrm{<span\; class="notranslate">\; \&delta;</span>}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 3</span>}<span\; class="notranslate">\; )</span>$

Among them, P _exp is the expected power value of the feed motor preset according to the actual processing, and δ is the power deviation rate. Calculate the power deviation rate between the preset power and the predicted power. If the deviation rate is less than the set value, there is no need to adjust the speed and continue power prediction; The feed rate is adjusted.

Step 3: Predict the value of the back knife, and realize the speed regulation through the preset power value and the back knife:

According to the power value predicted in step 1, as well as the system power equation and the empirical calculation formula of feed force, the functional relationship between the input power of the motor, the feed speed and the amount of back cutting is obtained, and then the amount of back cutting under actual working conditions is predicted.

The feed force F _x is the component force that has a greater influence on the feed system of the CNC machine tool. After the machine tool runs stably, the force acting on the workbench is equal to the feed force. According to the empirical calculation formula for the feed force in engineering, that is In the same way, the main cutting force F _z can be calculated, and the power consumed during cutting can be obtained at the same time P _m = F _z v; the shaft speed of the feed motor ω _m = K _g ω _ls , where ω _ls is the speed of the ball screw. The relationship between the table feed speed v and the ball screw speed is cutting speed Where ω is the spindle speed, and d is the spindle diameter. is the material coefficient; is the force correction coefficient; Correct the index for the back cut amount; Correct the index for the feed rate; is the cutting speed correction index; the relationship between the table feed speed and the feed rate is Substituting the above formulas into formula (2), it is possible to predict the counter-cutting amount a _p of the motor input power P _ax and the feed speed v, namely:

$\begin{array}{c}{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 3</span>}{{\mathrm{<span\; class="notranslate">\; RB</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; K</span>}}_{\mathrm{<span\; class="notranslate">\; T</span>}}{\mathrm{<span\; class="notranslate">\; B</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; )</span>{<span\; class="notranslate">\; (</span>\frac{\mathrm{<span\; class="notranslate">\; 2</span>}{\mathrm{<span\; class="notranslate">\; \&pi;K</span>}}_{\mathrm{<span\; class="notranslate">\; g</span>}}\mathrm{<span\; class="notranslate">\; v</span>}}{\mathrm{<span\; class="notranslate">\; P</span>}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 6</span>}{\mathrm{<span\; class="notranslate">\; RB</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; K</span>}}_{\mathrm{<span\; class="notranslate">\; T</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&CenterDot;</span><span\; class="notranslate">\; \&lsqb;</span>{\mathrm{<span\; class="notranslate">\; K</span>}}_{\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; q</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; m</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; m</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}\mathrm{<span\; class="notranslate">\; o</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; d</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; T</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}^{<span\; class="notranslate">\; \&prime;</span>}\\ <span\; class="notranslate">\; +</span>\frac{\mathrm{<span\; class="notranslate">\; 9.81</span>}{\mathrm{<span\; class="notranslate">\; PC</span>}}_{{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}}}{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; p</span>}}^{{\mathrm{<span\; class="notranslate">\; x</span>}}_{{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}}}}{<span\; class="notranslate">\; (</span>\frac{\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; \&pi;</span>}}{\mathrm{<span\; class="notranslate">\; \&omega;</span>}}\mathrm{<span\; class="notranslate">\; v</span>}<span\; class="notranslate">\; )</span>}^{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}}}}{<span\; class="notranslate">\; (</span>\frac{\mathrm{<span\; class="notranslate">\; \&omega;</span>}\mathrm{<span\; class="notranslate">\; d</span>}}{\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; \&pi;</span>}}<span\; class="notranslate">\; )</span>}^{{\mathrm{<span\; class="notranslate">\; no</span>}}_{{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}}}}{\mathrm{<span\; class="notranslate">\; K</span>}}_{{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}}\mathrm{<span\; class="notranslate">\; x</span>}}}{{\mathrm{<span\; class="notranslate">\; K</span>}}_{\mathrm{<span\; class="notranslate">\; T</span>}}\mathrm{<span\; class="notranslate">\; 2</span>}{\mathrm{<span\; class="notranslate">\; \&pi;K</span>}}_{\mathrm{<span\; class="notranslate">\; g</span>}}}<span\; class="notranslate">\; \&rsqb;</span>\frac{\mathrm{<span\; class="notranslate">\; 2</span>}{\mathrm{<span\; class="notranslate">\; \&pi;K</span>}}_{\mathrm{<span\; class="notranslate">\; g</span>}}\mathrm{<span\; class="notranslate">\; v</span>}}{\mathrm{<span\; class="notranslate">\; P</span>}}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 3</span>}\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; \&lsqb;</span>{\mathrm{<span\; class="notranslate">\; K</span>}}_{\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; q</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; m</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; m</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}\mathrm{<span\; class="notranslate">\; o</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; d</span>}}<span\; class="notranslate">\; )</span>\\ <span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; T</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; +</span>\frac{\mathrm{<span\; class="notranslate">\; 9.81</span>}{\mathrm{<span\; class="notranslate">\; PC</span>}}_{{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}}}{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; p</span>}}^{{\mathrm{<span\; class="notranslate">\; x</span>}}_{{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}}}}{<span\; class="notranslate">\; (</span>\frac{\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; \&pi;</span>}}{\mathrm{<span\; class="notranslate">\; \&omega;</span>}}\mathrm{<span\; class="notranslate">\; v</span>}<span\; class="notranslate">\; )</span>}^{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}}}}{<span\; class="notranslate">\; (</span>\frac{\mathrm{<span\; class="notranslate">\; \&omega;</span>}\mathrm{<span\; class="notranslate">\; d</span>}}{\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; \&pi;</span>}}<span\; class="notranslate">\; )</span>}^{{\mathrm{<span\; class="notranslate">\; no</span>}}_{{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}}}}{\mathrm{<span\; class="notranslate">\; K</span>}}_{{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}}\mathrm{<span\; class="notranslate">\; x</span>}}}{{\mathrm{<span\; class="notranslate">\; K</span>}}_{\mathrm{<span\; class="notranslate">\; T</span>}}\mathrm{<span\; class="notranslate">\; 2</span>}{\mathrm{<span\; class="notranslate">\; \&pi;K</span>}}_{\mathrm{<span\; class="notranslate">\; g</span>}}}{<span\; class="notranslate">\; \&rsqb;</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}\end{array}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 4</span>}<span\; class="notranslate">\; )</span>$

(4) can be written as the following form:

P _ax ＝Av ² +Bv+Cva _p +D(a _p ) ² +Ea _p +H (5)

When the working conditions change, the amount of back engagement a _p will change accordingly, and the feed speed v needs to be adjusted. The amount of cutting can be predicted by predicting the power value, and the predicted value of cutting amount can be brought back to the formula (5). According to the actual preset power, the feed speed that needs to be adjusted can be obtained, and then the speed regulation process of the system can be completed. .

_Pax can be kept constant with the change of a _p .

The beneficial effects of the present invention are:

This method establishes the functional relationship between the power and the back cutting amount, and uses the power of the main drive system and the feed system during the rough turning process of the machine tool to predict the back cutting amount in real time, so that it can be optimized in real time according to the actual cutting conditions. Feed speed, realize the control of feed speed. This method not only changes the machining method of cutting with a constant feed speed in the rough turning process of CNC machine tools, but also optimizes the feed speed in real time according to the actual cutting conditions, and effectively solves the problem of speed regulation lag, greatly improving Improve machine tool utilization and cutting efficiency.

Description of drawings

In order to make the purpose, technical scheme and beneficial effect of the present invention clearer, the present invention provides the following drawings for illustration:

Fig. 1 is a schematic flow sheet of the method of the present invention;

Figure 2 is a schematic diagram of cutting force and cutting parameters;

Fig. 3 is the comparative figure of cutting process power measured curve and predicted curve in the embodiment;

Fig. 4 is the simulation result of speed control in the embodiment;

detailed description

The preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

Fig. 1 is the schematic flow chart of the method of the present invention, as shown in the figure, the feed system speed regulation method in the rough turning process of the CNC machine tool based on power prediction according to the present invention comprises the following steps: Step 1: Predict the next moment Power value; Step 2: Calculate the deviation rate between the predicted power and the preset power, and judge whether speed regulation is required; Step 3: Control the speed of the feed system according to the preset power and the amount of back cutting.

Predict the power value at the next moment:

A voltage sensor, a current sensor and a power sensor are respectively installed to collect the voltage, current and power values at the input end of the feed motor. Filter the collected data and process the singular values; then calculate the collected voltage and current values, and fuse the calculated power value with the detected power value to improve the power detection accuracy.

The power flow of the servo motor in the feed system includes stator copper loss, iron loss, mechanical loss, stray loss, and the output power of the driving load. Combined with the steady-state voltage equation of the AC servo motor, the power of the servo motor can be expressed as:

${\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 3</span>}\mathrm{<span\; class="notranslate">\; R</span>}{\mathrm{<span\; class="notranslate">\; I</span>}}_{\mathrm{<span\; class="notranslate">\; the\; s</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span>\frac{{\mathrm{<span\; class="notranslate">\; \&omega;</span>}}_{\mathrm{<span\; class="notranslate">\; e</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&psi;</span>}}_{\mathrm{<span\; class="notranslate">\; d</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; \&psi;</span>}}_{\mathrm{<span\; class="notranslate">\; q</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; R</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; \&omega;</span>}}_{\mathrm{<span\; class="notranslate">\; e</span>}}{\mathrm{<span\; class="notranslate">\; K</span>}}_{\mathrm{<span\; class="notranslate">\; e</span>}}{\mathrm{<span\; class="notranslate">\; i</span>}}_{\mathrm{<span\; class="notranslate">\; q</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>$

Among them, R is the stator winding resistance; I _s is the effective value of the stator phase current; _{ω e} is the angular velocity of the electromagnetic field of the motor; ψ _d is the direct axis component of the magnetic flux; ψ _q is the quadrature axis component of the magnetic flux; K _e is the electromagnetic torque coefficient; is the quadrature axis component of the stator current. Combined with engineering practice, the power equation of the feed system can be obtained from formula (1):

in, K _T is the torque coefficient; B _m is the motor damping coefficient; P is the pitch of the ball screw; μ _v is the viscous friction coefficient; K _g is the transmission ratio of the coupling; ω _m is the shaft speed of the feed motor; μ _c is Coulomb friction coefficient; M _t is the mass of the workbench; M _load is the mass of the workpiece loaded on the workbench; T ₀ is the internal loss torque of the motor; F _ext is the force on the workbench.

Take the processed power value as input, construct the training sample (x _k , y), input it into the neural network, and obtain the output value y, which is the predicted value of the power at time k+1. The power prediction refers to estimating the power value P _m+h (h>0) at the time t _m+h based on the detected power data (P _m , P _m-1 , . . . ) before the time t _m . The theoretical basis for realizing this process is to perform nonlinear function fitting on historical data (P _m ,P _m-1 ,…,) to generate a time-varying function P _m (t) about variable t, and P _m+1 There is a mapping relationship with the function P _m (t), that is, P _m+1 =F[P _m (t)]. Therefore, by fitting the functional F[·], the value of P _m+h (h>0) can be estimated. This method uses a three-layer neural network with a single hidden layer to predict the power value, and the selected neural network model is:

Among them, v _i1 is the connection weight of the i-th hidden layer neuron to the output, n is the number of data collected at a certain time, K is the number of time series before k+1 time, ω _ki is the input neuron at the k-th time The connection weight of the unit to the i-th hidden layer neuron, x _k is the input at the k-th moment, and θ is the threshold value of the hidden layer neuron.

Calculate the deviation rate between the predicted power and the preset power to determine whether speed regulation is required:

By obtaining the predicted power value, the speed regulation equation of the feed system in the rough turning process of CNC machine tools based on power prediction can be obtained as:

$\frac{{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; exp</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; x</span>}}}{{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; exp</span>}}}<span\; class="notranslate">\; \&le;</span>\mathrm{<span\; class="notranslate">\; \&delta;</span>}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 3</span>}<span\; class="notranslate">\; )</span>$

Among them, P _exp is the expected power value of the feed motor preset according to the actual processing, and δ is the power deviation rate. Calculate the power deviation rate between the preset power and the predicted power. If the deviation rate is less than the set value, there is no need to adjust the speed and continue power prediction; The feed rate is adjusted.

Predict the value of the back knife, and realize speed regulation through the preset power value and the back knife

According to the power value predicted in step 1, as well as the system power equation and the empirical calculation formula of feed force, the functional relationship between the input power of the motor, the feed speed and the back cutting amount is obtained, and then the back cutting amount under actual working conditions is predicted .

Figure 2 is a schematic diagram of cutting force and cutting parameters. Under the action of the rake face of the tool, the metal is squeezed to generate cutting force. The cutting force can be decomposed into three mutually perpendicular cutting components: (1) cutting force F _z : the orthographic projection of the total cutting force in the main motion direction; (2) feed force F _x : the total cutting force in the feed direction Orthographic projection of ; (3) Backward force F _y : the component force of the total cutting force on the vertical working plane. (4) Back cutting a _p : the vertical distance between the machined surface and the surface to be machined; (5) Feed rate f: the displacement of the tool relative to the workpiece in the direction of feed movement; cutting speed v _c : The instantaneous velocity of a selected point on the cutting edge relative to the main motion of the workpiece.

The feed force F _x is the component force that has a greater influence on the feed system of the CNC machine tool. After the machine tool runs stably, the force acting on the workbench is equal to the feed force. According to the empirical calculation formula for the feed force in engineering, that is In the same way, the main cutting force F _z can be calculated, and the power consumed during cutting can be obtained at the same time P _m = F _z v; the shaft speed of the feed motor ω _m = K _g ω _ls , where ω _ls is the speed of the ball screw. The relationship between the table feed speed v and the ball screw speed is cutting speed Where ω is the spindle speed, d is the spindle diameter; is the material coefficient; is the force correction coefficient; Correct the index for the back cut amount; Correct the index for the feed rate; The relationship between the feed rate and the feed rate of the exponential worktable is corrected for the cutting speed as Substituting the above formulas into formula (2), it is possible to predict the counter-cutting amount a _p of the motor input power P _ax and the feed speed v, namely:

$\begin{array}{c}{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 3</span>}{{\mathrm{<span\; class="notranslate">\; RB</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}}^{<span\; class="notranslate">\; \&prime;</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; K</span>}}_{\mathrm{<span\; class="notranslate">\; T</span>}}{\mathrm{<span\; class="notranslate">\; B</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; )</span>{<span\; class="notranslate">\; (</span>\frac{\mathrm{<span\; class="notranslate">\; 2</span>}{\mathrm{<span\; class="notranslate">\; \&pi;K</span>}}_{\mathrm{<span\; class="notranslate">\; g</span>}}\mathrm{<span\; class="notranslate">\; v</span>}}{\mathrm{<span\; class="notranslate">\; P</span>}}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 6</span>}{\mathrm{<span\; class="notranslate">\; RB</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; K</span>}}_{\mathrm{<span\; class="notranslate">\; T</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&Center\; Dot;</span><span\; class="notranslate">\; \&lsqb;</span>{\mathrm{<span\; class="notranslate">\; K</span>}}_{\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; q</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; m</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; m</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}\mathrm{<span\; class="notranslate">\; o</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; d</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; T</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}^{<span\; class="notranslate">\; \&prime;</span>}\\ <span\; class="notranslate">\; +</span>\frac{\mathrm{<span\; class="notranslate">\; 9.81</span>}{\mathrm{<span\; class="notranslate">\; PC</span>}}_{{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}}}{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; p</span>}}^{{\mathrm{<span\; class="notranslate">\; x</span>}}_{{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}}}}{<span\; class="notranslate">\; (</span>\frac{\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; \&pi;</span>}}{\mathrm{<span\; class="notranslate">\; \&omega;</span>}}\mathrm{<span\; class="notranslate">\; v</span>}<span\; class="notranslate">\; )</span>}^{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}}}}{<span\; class="notranslate">\; (</span>\frac{\mathrm{<span\; class="notranslate">\; \&omega;</span>}\mathrm{<span\; class="notranslate">\; d</span>}}{\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; \&pi;</span>}}<span\; class="notranslate">\; )</span>}^{{\mathrm{<span\; class="notranslate">\; no</span>}}_{{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}}}}{\mathrm{<span\; class="notranslate">\; K</span>}}_{{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}}\mathrm{<span\; class="notranslate">\; x</span>}}}{{\mathrm{<span\; class="notranslate">\; K</span>}}_{\mathrm{<span\; class="notranslate">\; T</span>}}\mathrm{<span\; class="notranslate">\; 2</span>}{\mathrm{<span\; class="notranslate">\; \&pi;K</span>}}_{\mathrm{<span\; class="notranslate">\; g</span>}}}<span\; class="notranslate">\; \&rsqb;</span>\frac{\mathrm{<span\; class="notranslate">\; 2</span>}{\mathrm{<span\; class="notranslate">\; \&pi;K</span>}}_{\mathrm{<span\; class="notranslate">\; g</span>}}\mathrm{<span\; class="notranslate">\; v</span>}}{\mathrm{<span\; class="notranslate">\; P</span>}}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 3</span>}\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; \&lsqb;</span>{\mathrm{<span\; class="notranslate">\; K</span>}}_{\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; q</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; m</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; m</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}\mathrm{<span\; class="notranslate">\; o</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; d</span>}}<span\; class="notranslate">\; )</span>\\ <span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; T</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; +</span>\frac{\mathrm{<span\; class="notranslate">\; 9.81</span>}{\mathrm{<span\; class="notranslate">\; PC</span>}}_{{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}}}{\mathrm{<span\; class="notranslate">\; a</span>}}_{\mathrm{<span\; class="notranslate">\; p</span>}}^{{\mathrm{<span\; class="notranslate">\; x</span>}}_{{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}}}}{<span\; class="notranslate">\; (</span>\frac{\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; \&pi;</span>}}{\mathrm{<span\; class="notranslate">\; \&omega;</span>}}\mathrm{<span\; class="notranslate">\; v</span>}<span\; class="notranslate">\; )</span>}^{{\mathrm{<span\; class="notranslate">\; the\; y</span>}}_{{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}}}}{<span\; class="notranslate">\; (</span>\frac{\mathrm{<span\; class="notranslate">\; \&omega;</span>}\mathrm{<span\; class="notranslate">\; d</span>}}{\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; \&pi;</span>}}<span\; class="notranslate">\; )</span>}^{{\mathrm{<span\; class="notranslate">\; no</span>}}_{{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}}}}{\mathrm{<span\; class="notranslate">\; K</span>}}_{{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}}\mathrm{<span\; class="notranslate">\; x</span>}}}{{\mathrm{<span\; class="notranslate">\; K</span>}}_{\mathrm{<span\; class="notranslate">\; T</span>}}\mathrm{<span\; class="notranslate">\; 2</span>}{\mathrm{<span\; class="notranslate">\; \&pi;K</span>}}_{\mathrm{<span\; class="notranslate">\; g</span>}}}{<span\; class="notranslate">\; \&rsqb;</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}\end{array}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 4</span>}<span\; class="notranslate">\; )</span>$

Speed regulation through power value and back cutting amount

(4) can be written as the following form:

P _ax ＝Av ² +Bv+Cva _p +D(a _p ) ² +Ea _p +H (5)

When the working conditions change, the amount of back engagement a _p will change accordingly, and the feed speed v needs to be adjusted. The amount of cutting can be predicted by predicting the power value, and the predicted value of cutting amount can be brought back to the formula (5). According to the actual preset power, the feed speed that needs to be adjusted can be obtained, and then the speed regulation process of the system can be completed. .

Example:

In this embodiment, the feasibility of the proposed feed speed control method is verified by a method combining experiment and simulation. When formulating the verification scheme, because the control interface of the CNC machine tool used in the experiment could not be obtained, the verification process was divided into two parts. First, the feasibility of the power prediction method is verified by testing the cutting process of the CNC machine tool; then, the static simulation analysis of the feed speed control is carried out with the measured power value as the data source.

(1) Power prediction test

The Z-axis feed system of a CNC lathe (C2-6136HK/1) was used for the test. The basic parameters of the feed system of the machine tool are shown in Table 1, and the test conditions are shown in Table 2.

Table 1 Power related parameters (Z axis) of feed system of CNC lathe C26136HK/1

Table 2 Test conditions of feed system of CNC machine tools

After the machine tool runs stably, the comparison between the measured power curve and the predicted power curve in the cutting process is shown in Figure 3, where the power sampling period is about 40 ms. It can be seen from the comparison chart that the predicted curve has a high fitting degree to the measured curve, which verifies the feasibility of the power prediction method.

(2) Speed control simulation

The basic process of speed control is as follows:

Assume that the machine tool runs stably at time t ₀ and the cutting process starts. At this time v=v ₀ , a _p =a _p0 , P _ax =P ₀ , and in the interval [t ₀ -t _i ) satisfies In this interval, the controller does not act, and the feed rate remains unchanged at _v0 .

time t _i At this time, it shows that a _pi changes, and the power deviation rate is greater than the set value δ. At this point, the system will calculate a _pi based on P _i , v ₀ . Then calculate v _i according to P _exp and a _pi . At the same time, the speed controller acts to make the feed speed reach v _i .

At the beginning of controller design, both acceleration and deceleration are controlled by PID. Judging from the simulation curve, the control effect is very good. However, through the analysis of the speed change curve, we found that the existence of the differential link shortens the response time during acceleration and speed regulation. In the actual processing process, if the speed change is too severe during the acceleration process, shock vibration is likely to occur. Therefore, the control method is revised, the acceleration is changed to PI control, and the deceleration is controlled by PID. The speed control simulation results are shown in Figure 4. The proposed control model can realize the speed control, the speed change is relatively gentle, and the power value is basically stable at the set value.

Finally, it should be noted that the above preferred embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail through the above preferred embodiments, those skilled in the art should understand that it can be described in terms of form and Various changes may be made in the details without departing from the scope of the invention defined by the claims.

Claims (1)

1. a feed system speed regulation method in the rough turning process of numerical control machine tool based on power prediction, it is characterized in that: comprise the following steps: Step 1: Predict the power value at the next moment: A voltage sensor, a current sensor and a power sensor are respectively installed to collect the voltage, current and power values at the input end of the feed motor. Filter the collected data separately, and process the singular values. Take the processed power value as input, construct the training sample (x _k , y), input it into the neural network, and obtain the output value y, which is the predicted value of the power at time k+1. This method uses a three-layer neural network with a single hidden layer to predict the power value, and the selected neural network model is: Among them, v _i1 is the connection weight of the i-th hidden layer neuron to the output, n is the number of data collected at a certain time, K is the number of time series before k+1 time, ω _ki is the input neuron at the k-th time The connection weight of the unit to the i-th hidden layer neuron, x _k is the input at the k-th moment, and θ is the threshold value of the hidden layer neuron. Step 2: Calculate the deviation rate between the predicted power and the preset power, and judge whether speed regulation is required: By obtaining the predicted power value, the speed regulation equation of the feed system in the rough turning process of CNC machine tools based on power prediction is obtained as: $\frac{{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; exp</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; x</span>}}}{{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; exp</span>}}}<span\; class="notranslate">\; \&le;</span>\mathrm{<span\; class="notranslate">\; \&delta;</span>}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>$ Among them, P _exp is the expected power value of the feed motor preset according to the actual processing, and δ is the power deviation rate. Calculate the power deviation rate between the preset power and the predicted power. If the deviation rate is less than the set value, then there is no need to adjust the speed and continue power prediction; Adjust for speed. Step 3: Predict the value of the back knife, and realize the speed regulation through the preset power value and the back knife: According to the power value predicted in step 1, as well as the system power equation and the empirical calculation formula of feed force, the prediction function of motor input power, feed speed and back cutting amount can be obtained as shown in formula (2). The motor input power P _ax and There is a quadratic function relationship between the feed speed v and the amount a _p of the knife. P _ax ＝Av ² +Bv+Cva _p +D(a _p ) ² +Ea _p +H (2) Among them, A, B, C, D, E are the correlation coefficients of the quadratic function, and H is a constant. Through formula (2), the quadratic function relationship between the predicted amount of back cutting and feed speed can be obtained. Under the premise of keeping the power constant, when the working condition changes, the amount of back engagement a _p will change accordingly. Adjusting the feed speed v can keep _Pax constant with the change of a _p .